HN user

cantaloupe

565 karma
Posts14
Comments64
View on HN

I didn’t realize that GPT-5.6 is basically dominating the cost/intelligence Pareto frontier right now, at least for this set of benchmarks. Otherwise it’s only Fable on the very high end and DeepSeek on the very low end. This Kimi model gets close, though.

I browsed the GitHub and website for a bit, but didn’t see any examples! It would be useful to share the output for a common question that can be substantiated with reliable sources, like “Is coffee good for me?”. Even better if you can show a comparison to other deep research tools. From the copy it seems like Cleverbee could pull from more diverse sources (e.g. YouTube) and fewer unreliable sources (e.g. product blogs). Show that off!

His paper is #55, “Some upsetting things about shapes”, found on page 342. Its header and contents are not searchable, presumably due to his use of his bespoke typesetting program from a previous SIGBOVIK year. It’s a great read!

NPR is not a monolithic media organization. In my experience, local NPR stations are one of the best sources of interesting and relevant local news. In contrast, most local TV/Radio news is borderline a crime blotter ginned up to keep people outraged.

Regarding the national NPR newsroom, I think this story will provoke positive change, as indicated in the article. There is no media which every person would consider unbiased, and very few media organizations take action to even attempt to reign in biases. The fact that editors will start reviewing coverage more closely to remove tilt sets a higher bar than all but a few news organizations.

I chuckle thinking about a reporter stepping out of another random news room in the country and spreading outrage that the coverage has a bias. The response would generally be: “Yes, duh.”

2023 Hugo Awards 3 years ago

I haven’t read The Children of Time, but I would highly recommend Tchaikovsky’s Final Architecture trilogy. It’s been my favorite series I’ve read in a few years.

I’m far from a financial analyst myself, but I’m pretty sure the people who come up with the “expected” numbers are financial analysts who would be well aware of Tesla’s public pricing changes. So a better way of looking at is that people with a vested interest in correctly estimating earnings based on all available knowledge were surprised to the downside.

Is this satire?

- dubious IQ (dIQ) measuring

- incorrect accusation of “putting them down” immediately following an ad hominem attack

- unspecific “insight” that EVs sacrifice practical things to get distance, when the parent and OP were both basically saying the opposite: EVs sacrifice distance for practicality

In a walkable city, gyms and grocery stores are walkable from where you live, so there is no need to string together a bunch of stops. When you can walk to the gym or the grocery store in less than 10 minutes, no pre-planning is needed to optimize a route. And one usually doesn’t need more than a backpack full of groceries if within walking distance of a grocery store, because stocking up isn’t necessary. It’s amazing how much a truly walkable area simplifies life.

That’s fair, although people often underestimate the time “more convenient” options take. For example, walk 2 minutes to a bus, wait 5 minutes, ride 5 minutes, walk 3 minutes to destination. Not too much faster. Happens all the time with cars in the city. Spend 5 minutes looking for parking that ends up being a 5 minute walk away, and a 10 drive is now a 20 minute transit.